U.S. patent number 8,260,777 [Application Number 11/223,809] was granted by the patent office on 2012-09-04 for server system and methods for matching listings to web pages and users.
This patent grant is currently assigned to A9.com, Inc.. Invention is credited to Udi Manber.
United States Patent |
8,260,777 |
Manber |
September 4, 2012 |
**Please see images for:
( Certificate of Correction ) ** |
Server system and methods for matching listings to web pages and
users
Abstract
A listing creation application provides functionality for a
listing provider to create a listing to be displayed on web pages,
and to associate this listing with a content descriptor that
specifies the type of content to which the listing is to be
matched. The content descriptor may, for example, be a topic
selected from a hierarchy of topics, a keyword, or a keyword
phrase. Upon selecting a particular content descriptor, the listing
provider is presented with a corresponding list of user classes
with which the listing may be associated. A real time matching
engine uses a listing's associations with specific content
descriptors and user classes to match the listing to web page
requests, such that the listings selected for presentation to a
user are dependent upon both the profile of the user and the type
of content being requested.
Inventors: |
Manber; Udi (Palo Alto,
CA) |
Assignee: |
A9.com, Inc. (Palo Alto,
CA)
|
Family
ID: |
46726588 |
Appl.
No.: |
11/223,809 |
Filed: |
September 9, 2005 |
Current U.S.
Class: |
707/731; 715/962;
705/14.1 |
Current CPC
Class: |
G06F
16/957 (20190101) |
Current International
Class: |
G06F
17/30 (20060101) |
Field of
Search: |
;273/277 ;715/962
;705/14 |
References Cited
[Referenced By]
U.S. Patent Documents
Other References
Gibbs, "The EBay Seller's Guide", May 1, 2005,
www.auction-revolution.com/sellerguide.shtml, retrieved Nov. 6,
2007, p. 1-18. cited by examiner .
Karp, "eBay Hacks: 100 Industrial-Strength Tips & Tools",
O'Reilly Press, p. 7, 19, 26,28,32,34,35,38,44,45,49,56,78,85-88,
102, 110-115,123, 148, 174, 195-198, 244-245,264-265, 310. cited by
examiner .
Karp, "EBay Hacks: 100 Industrial-Stringth Tips & Tools", Jun.
2005, O'Reilly Press, 2nd Edition, Hack #12, Repaginated numbering
of p. 1-3 (approx. mapping from pp. 28-31). <Retrieved from
Safari Books Dec. 22, 2008>, hereinafter Karp2. cited by
examiner .
Pavlov et al, "Supporting Privacy in Decentralized Additive
Reputation", Mar. 2004, 2nd International Conf. on Trust Mgt., p.
1-11, <Retrieved from
http://www.cs.huji.ac.i1/.about.jeff/papers.html May 21, 2009>.
cited by examiner .
Motani et al, "Peoplenet: engineering a wireless virtual social
network", Aug. 2005, MobiCom'05, ACM Press, 15 pages, <Retrieved
from CiteSeer May 21, 2009>. cited by examiner .
Spring et al, "3 Minutes: The Man Behind Craigslist", Nov. 2004, PC
World, p. 32, <Retrieved from PC World May 21, 2009>. cited
by examiner .
"Business: The online ad attack; Internet advertising." The
Economist Apr. 30, 2005: 63. ABI/INFORM Global. ProQuest, May 21,
2009 <http://www.proquest.com/ on May 21, 2009>. cited by
examiner.
|
Primary Examiner: Pham; Khanh
Attorney, Agent or Firm: Knobbe, Martens, Olson & Bear
LLP
Claims
What is claimed is:
1. A computer-implemented method of generating web pages that
include listings selected from a database, the method comprising:
providing a database that stores listings submitted by listing
providers via a user interface of a listing creation application,
said listings stored in said database in association with specific
content descriptors and target user classes specified by
corresponding listing providers via said listing creation
application, said listings including a first listing submitted by a
first listing provider, wherein the content descriptors include
keywords and/or topics for matching the listings to web pages, and
at least some of the user classes correspond to different content
descriptors than other user classes; receiving a request for a web
page from a user other than the first listing provider, and
identifying a content descriptor associated with the request and/or
the web page; identifying a plurality of user classes associated
with the content descriptor, and selecting, from the plurality of
user classes, a user class that characterizes the user; selecting
at least one listing that is stored in the database in association
with the both the content descriptor and the user class, said at
least one listing including said first listing submitted by the
first listing provider, said user class corresponding uniquely to
the content descriptor, and consisting of users that have selected
a specified type of catalog item associated with the content
descriptor; and incorporating the at least one listing into the web
page for display to the user; wherein the method is performed by a
computerized system that includes one or more physical servers.
2. The method of claim 1, wherein the content descriptors include
topics arranged within a hierarchy of topics.
3. The method of claim 2, wherein the step of identifying a content
descriptor associated with the request and/or the web page
comprises identifying a topic of the web page.
4. The method of claim 1, wherein the request is a search request
that includes a search query, and the step of identifying a content
descriptor associated with the request and/or the web page
comprises identifying a content descriptor associated with the
search query.
5. The method of claim 1, wherein the content descriptors include
keywords.
6. The method of claim 5, wherein the request is a search request
that includes a search query, and the step of identifying a content
descriptor associated with the request and/or the web page
comprises identifying a keyword included within the search
query.
7. The method of claim 1, wherein the first listing is selectable
by the user on the web page to access a web site of said first
listing provider.
8. The method of claim 1, further comprising providing a web-based
listing creation application that provides a user interface and
functionality for a listing provider to specify a listing to add to
the database, associate the listing with a selected content
descriptor, view a list of user classes associated with the
selected content descriptor, and select from the list of user
classes to specify a target user class.
9. The method of claim 1, further comprising responding to user
selection of the first listing, as displayed on said web page, by
causing a monetary amount to be charged to the first listing
provider.
Description
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to user interfaces and document
processing methods for controlling the selection of listings from a
database to display on web pages.
2. Description of the Related Art
Various types of systems exist for displaying "paid placement" type
listings on web pages. One type of system allows listing providers
to associate their respective listings with specific keywords. When
a user submits a search query to an Internet search engine, the
listing or listings associated with the keyword or keywords entered
by the user are displayed on the resulting search results
page--typically under a heading such as "sponsored links" or
"sponsored results." The display position of each such listing, and
the amount paid by the corresponding listing provider if the
listing is clicked on by a user, are typically dictated by an
associated bid amount specified by the listing provider. Typically,
the listings are displayed on a variety of different independent
web sites through a process known as syndication.
In some systems of the type described above, the listing providers
can additionally specify target geographic regions in which their
listings are to be displayed. For example, a merchant that does
business in a particular geographic region may, in addition to
associating its listing with one or more keywords, request that the
listing be displayed to users in that region. In such systems, the
IP (Internet Protocol) addresses associated with the search
requests are used to identify or predict the locations of the
users.
Another type of system operates by selecting paid placement
listings to display on relatively static web pages by performing
content-based analyses of such pages. For example, Google's AdSense
program allows web site operators to configure selected pages of
their respective web sites to display or "host" listings served by
a remote listing server. To select appropriate listings to display
on a given page, a crawler program analyzes the textual content of
the page, and attempts to match the page's content to relevant
listings.
A significant problem with these systems is that they typically
select paid placement listings to display on web pages without
taking into consideration information about the preferences or
prior actions of the user who is requesting the page. As a result,
the listings are frequently of little or no value to the target
user.
SUMMARY OF THE INVENTION
The present invention comprises a listing generation and matching
system that enables listing providers to more accurately and
precisely target listings to web users. The invention may, for
example, be implemented in conjunction with an electronic catalog
system that maintains profiles (browse histories, purchase
histories, rental histories, histories of searches performed,
product reviews submitted, etc.) of users of an electronic catalog
of items.
In one embodiment, the listing generation and matching system
includes a listing creation application that provides functionality
for a listing provider to create a listing, and to associate this
listing with a content descriptor that specifies the type(s) of web
page content to which the listing is to be matched. The content
descriptor may, for example, be a topic selected from a browsable
hierarchy of topics, a keyword, or a keyword phrase. Upon selecting
a particular content descriptor, the listing provider is also
presented with a corresponding list of target user classes or
attributes with which to associate the listing. For example, if the
content descriptor "digital photography" is selected, the listing
provider may be given the option to associate the listing with one
or more of the following user classes: "recent purchasers of a
digital camera," "recent purchasers of digital cameras having a
ComactFLASH slot," "users who have recently viewed a digital camera
without making a purchase," "users who purchased a photo printer,
photo paper, or photo printer ink in last N days" "users who have a
digital camera in their respective shopping carts," "users who
viewed a product detail page of a digital camera during the current
browsing session," "all users."
Listings created by listing providers via the listing creation
application are stored in a database in association with the
designated content descriptors and user classes. A real time
matching engine uses these associations to dynamically select
listings from the database to display on web pages. For example,
when a user requests a particular web page, the real time matching
engine may attempt to locate one or more listings that are
associated with both (a) a content descriptor that matches the
content being requested by the user, and (b) an associated user
class of which the user is a member. For instance, if a user who
recently purchased a digital camera requests a catalog page, or
submits a search query, related to digital photography, the
matching engine may select a listing that is associated with both
the content descriptor "digital photography" and the user class
"recent purchasers of a digital camera."
The user profile data used by the real-time matching engine may
come from any source or combination of sources, including but not
limited to web sites that collect user-specific information about
the browsing activities of users. Listings selected by the matching
engine may be displayed exclusively on pages of a single web site,
such as a web site that maintains user profiles of its users, or
may be displayed across multiple distinct web sites using
syndication methods.
Neither this summary nor the following detailed description
purports to define the invention. The invention is defined by the
claims.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 illustrates a listing generation and matching system
according to one embodiment of the invention.
FIG. 2 illustrates an example web page of a user interface for
creating listings that are associated with specific topics and user
classes.
FIG. 3 illustrates a sequence of steps that may be performed by the
matching engine of FIG. 1 to select one or more listings to display
in response to a user request.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Specific embodiments of the invention will now be described with
reference to the drawings. These embodiments are provided in order
to illustrate, and not limit, the invention. The scope of the
invention is defined only by the claims.
I. OVERVIEW
FIG. 1 illustrates a listing generation and matching system 30
according to one embodiment of the invention. As illustrated, the
system 30 includes a web server system 32 that hosts a web-based
listing creation application 34. The listing creation application
34 provides a user interface and functionality for listing
providers, such as those wishing to advertise specific products or
services, to type in or upload paid placement or "sponsored link"
type listings (typically in the form of clickable advertisements),
and to associate or "link" these listings with specific topics and
user attributes. For instance, a company wishing to advertise its
online photo print service may define a clickable ad listing for
its service, and associate the listing with both the topic "digital
photography" and the user attribute "has recently purchased digital
camera."
A matching engine 38 uses these associations to match listings to
web pages and users in real time as pages are requested. For
instance, the matching engine 38 may select the online photo print
service listing for display primarily or exclusively when a user
who has recently purchased a digital camera requests a web page, or
conducts a search, related to digital photography. Because the
listings are selected based on the context of the user's browsing
session (e.g., the search being performed or the type of page being
accessed), and based further on attributes of the particular user,
the listings are more likely to be of interest to the users.
The listings are preferably selected by the matching engine 38 for
display on pages of a web site served by a second web server system
42, referred to herein as a web content server. The web content
server 42 hosts a web-based catalog application 44 that provides
browsable and searchable access to a database 46 of catalog items.
The catalog items stored or represented in the database 46 may, for
example, include physical products that can be purchased or rented,
digital products (journal articles, news articles, music files,
video files, software downloads, etc.) that can be purchased and/or
downloaded by users, web sites arranged by category within a web
site directory, subscriptions, and/or other types of items that may
be browsed and selected from a catalog.
The matching engine 38 may run on the physical server or servers of
the web content server 42, or may run on one or more separate
physical servers. In some embodiments, a single matching engine 38
may select ads to display on multiple distinct web sites hosted by
physically distinct web server systems. Well-known listing
syndication methods may be used for this purpose.
As depicted by the user computing devices 50 in FIG. 1, users may
access the web content server 42 and associated web site over the
Internet via a web browser to locate catalog items of interest. To
assist users in locating items, the catalog application 44 may
provide a browse tree in which the catalog items are arranged by
category, and/or a search engine through which users can submit
search queries to search for specific items. Depending on the
nature of the particular system, the catalog application 44 may
also include functionality for users to purchase, download, review,
rate, sell, and/or access specific catalog items.
As depicted in FIG. 1, information about the users is recorded by
the web content server 42 within a "user profiles" data repository
54. This data repository 54, which may include one or more
databases, stores a variety of different types of data about the
actions performed by particular users. The data stored for a given
user may include an "item selection history" indicative of the
catalog items selected by the user for purchasing, downloading,
viewing, rating, reviewing, renting, selling, adding to a shopping
cart, adding to a rental queue, accessing, and/or performing some
other type of affinity-evidencing action. Each item selection event
recorded in the user's item selection history may, for example,
include an event time/date stamp, an identifier of the selected
item and item category, and an identifier of the type of selection
action performed (e.g., purchase, view, shopping cart add, etc.).
Other types of profile information may also be stored in the user
profile data repository 54, including, for example, email
addresses, shipping addresses, search histories, browse category
selection histories, subscriptions, and memberships. As described
below, these and other types of user profile data may be used
directly or indirectly by the matching engine 38 to select listings
to display on web pages. Some of the profile data, and particularly
profile data associated with users' clickstreams, may be collected
and stored as described in US Patent Pub. 2005/0033803 and/or U.S.
Pat. No. 6,912,505, the disclosures of which are hereby
incorporated by reference.
With further reference to FIG. 1, the system 30 includes a
hierarchy 60 of subjects or "topics" with which listings may be
explicitly associated. The topics may, but need not, include
categories of items in the database 46, such as browse categories
of a browse tree. The following are examples of topics and
subtopics that may be included in the hierarchy: Electronics/DVD
Players; Financial Services/Tax Preparation Services; Web
Sites/Online Communities/Social Networking Sites. Many hundreds or
thousands of different topics may be provided within the hierarchy
60. Listing providers may browse or search this hierarchy 60 via
the listing creation application 34 to select specific topics with
which to associate their listings. Each topic serves generally as a
content category or descriptor for matching listings to web pages,
and may be represented in computer storage by a topic name and an
associated topic ID. As discussed below, the system 30 may
additionally or alternatively support the ability for listing
providers to associate their listings with specific keywords and
keyword phrases.
As further illustrated in FIG. 1, the system 30 also includes a
table or other structure 62 that specifies, for at least some of
the topics in the hierarchy 60 (or alternatively, for specific
keywords and keyword phrases), corresponding user "classes" that
may be selected by the listing provider to designate a target
audience. Each such class may be defined by a set of one or more
user attributes. Examples of classes that may be defined within the
system include the following:
users who have purchased a DVD player in last six months;
users who have downloaded at least one music file;
users with shipping addresses in San Francisco area;
users with email addresses associated with software companies;
users who have sold items via an electronic marketplace;
users who have searched for <keyword> during current browsing
session;
users who have birthdays within the next N weeks;
users who have purchased items for children;
users who have purchased a book on house remodeling;
users with an item falling in category C in their shopping carts or
on their wish lists;
users who are members of an online social networking service;
users with subscriptions to publication P;
users who have made at least one purchase.
As indicated by these example user classes, at least some of the
user classes are preferably unique to a particular topic, or subset
of related topics, within the topical hierarchy 60. In other cases,
the user classes may be generally applicable to all of the topics
within the hierarchy 60. A given topic may be associated with zero,
one, or multiple different user classes. The user classes that are
provided for a given topic may be specified manually by a system
administrator, and/or may be defined automatically via an automated
computer process that detects associations between particular user
attributes and particular topics. In addition to using predefined
user classes, the system may allow the listing provider to define
their own classes of target users. For example, the user interface
of the listing creation application 34 may display a list of user
attributes that may be used to define a new user class, and may
provide check boxes for selecting a set of user attributes to
use.
Listing providers may use the listing creation application 34 to
create new listings generally as follows. Initially, the listing
provider browses or searches the hierarchy 60 via a web browser to
locate a topic of interest. In some embodiments, the listing
providers may only be permitted to associate their listings with
topics falling at certain levels of the hierarchy 60, such as the
lowest (most specific) level or levels. Upon identifying a desired
topic, the listing provider can, for at least some topics, select
from a topic-specific list of user classes to which the listing is
to be targeted (see FIG. 2, discussed below). Selection of topics
and user classes may be performed by inclusion (e.g., associate ad
with "electronics/DVD Players/Owner of DVD Player") and/or by
exclusion (e.g., associate ad with all subtopics of "electronics"
except "televisions.")
The listing provider may also type in or upload the text or other
content of the listing. Typically, each listing includes a short
textual description (e.g., "Unlimited DVD Rentals for $19/Month")
and a link to an associated web site of the listing provider. Other
types of listings, including auction listings, classified ad
listings, and job listings, may also be supported.
The listing provider may also be permitted or required to specify
other listing parameters. For example, the listing provider may
specify an associated bid amount. The bid amount may be equal to,
or may be used to determine, the monetary amount that will be
charged to the listing provider each time the listing is clicked on
(selected) by a user. The bid amounts may also be used to rank
listings for display. Although bidding is preferably used, other
types of compensation schemes, including fixed pricing, may
alternatively be used. Other types of listing parameters that may
be specified by the listing provider include, for example, duration
of listing and maximum dollar amount to be charged per unit of
time.
The user interface for creating listings may also allow the
advertiser/listing provider to specify different bid amounts for
different user classes within a given topic. This may be useful
where, for example, the advertiser wishes to target multiple user
classes, but has different preferences for different user classes.
For instance, a target user who purchased a digital camera within
the last few months may be deemed more valuable to an advertiser of
camera accessories than a target user who simply browsed a digital
camera catalog page. Therefore, the advertiser may wish to bid
higher for one user class versus another user class.
FIG. 2 illustrates an example web page of the listing creation
application's user interface. In this example, the listing
provider/advertiser has navigated the topical hierarchy 60 to the
topic "digital photography," and has entered the text of a listing
(advertisement) to associate with this topical category. The web
page displays a set of check boxes that allow the listing provider
to optionally limit the display of the listing to one or more
classes of users. The listing provider can also specify a bid
amount for the listing. Although only a single bid field is
provided in this example, separate bid fields may alternatively be
provided for each of the user classes.
As further depicted in FIG. 1, listings generated by listing
providers via the listing creation application 34 are stored in a
listings database 70. Each listing may be stored in this database
70 in associate with identifiers of the topic(s) and user class(es)
with which the listing is associated, an identifier of the listing
provider, a bid amount (if bidding is used), and other listing
parameters such as those listed above. In some embodiments, a given
listing may be stored in association with multiple different
topics, and may be associated with different user classes within
each of the multiple topics.
The matching engine 38 accesses the listings database 70 to select
and retrieve listings to display on web pages served by the web
content server 42. The task of selecting listings is preferably
performed in real time in response to search query submissions
and/or non-search page requests from users. For example, when a
user requests a particular page of the catalog, the matching engine
38 may use both information about the requested page and
information about the user to select one or more listings to
display on the page. The information about the page may include its
content, its location within a directory structure, its
pre-assigned topical category, the browse category to which the
page corresponds, or any other information that may be used to
identify a topic associated with the page. In the case of a search
query submission, the matching engine 38 may use the search query
itself, and/or the results of the search query, to determine a
topic to which the user request corresponds.
The user information taken into consideration by the matching
engine 38 may include a history of catalog items viewed, purchased,
or otherwise selected, a history of browse categories accessed by
the user, a history of search queries submitted by the user, a
history of listings selected by the user, the user's shipping
address or email address, any combination of the foregoing, and/or
any other user profile information that may evidence the user's
preferences.
Although the task of analyzing the user profile information may be
performed in real time by the matching engine 38, performance may
be improved by periodically processing the relevant user profile
data, and identifying the user classes to which each user
corresponds, in an off-line mode. Thus, the user profile data
accessed by the matching engine 38 in response to a given page
request may include or consist of a preexisting list of user
classes of which the user is a member.
FIG. 3 illustrates a sequence of steps that may be performed by the
web content server 42 and matching engine 38 to respond to a search
query submission or a non-search page request from a user. As
depicted in blocks 90 and 92, if the request is a search request,
the topic most closely associated with the user's search query is
initially identified. This may be accomplished, for example, by
applying the search query to the textual descriptions of the topics
in hierarchy 60. General search queries may tend to map to higher
level nodes (topics) of the hierarchy 60, while more specific
search queries may tend to map to lower level nodes. As mentioned
above, the results of the search query may also be used to identify
the topic. In either case, the identified topic represents the
topic of the requested search results page.
As depicted by block 94, if the user's request is not a search
request, the topic is instead identified based on the identity
and/or content of the requested page. This may be accomplished by
analyzing the page's content in real time to identify the topic
most closely associated with the page. Alternatively, the topic may
be looked up from a previously generated mapping table that maps
pages to topics, or may be read from coding included in the
associated HTML document. The page's topic may also be determined
based on a location of the page in a directory structure.
Once the topic of the requested page has been identified, the
matching engine 38 uses the table 62 to identify the user classes,
if any, associated with this topic (block 96). As depicted in block
98, the matching engine 38 then uses profile data of the user to
select the user class (or set of user classes) that best
characterizes the user. As mentioned above, this task may involve
analyzing the user's raw profile data in real time, and/or may
involve looking up a previously generated list of the user classes
of which the user is a member. If the user is not a member of any
of the classes associated with the particular topic, or
insufficient user profile data exists to classify the user, one or
more listings may be selected based solely on the topic associated
with the user's request.
In block 100, the matching engine 38 retrieves from the listings
database 70 some or all of the listings associated with both the
topic and user class(es) (or just the topic if user class is not
applicable). These listings are incorporated into the web page
(which may be a search results page) requested by and returned to
the user (block 102). The order in which the listings are displayed
on the web page may depend on the associated bid amounts, listing
relevance scores, and/or other criteria.
As will be apparent from the foregoing, an important benefit of the
above-described process for creating and selecting listings is that
the advertiser, or other listing provider, can select a target
audience both in terms of the type of content being requested and
attributes of the users requesting such content. As a result, the
displayed listings tend to be highly relevant to the users'
interests. To further illustrate how the process may be used, the
following examples are provided:
Example 1
An online seller of Spanish-language PC training videos creates an
ad listing for its web site. The online seller also links this
listing to the topic "Computers," and to the user class "purchasers
of Spanish-language items." The listing will thereafter be
displayed primarily or exclusively on computer-related web pages,
and pages generated in response to computer-related searches, when
such pages are accessed by users who have purchased one or more
Spanish-language items from the catalog.
Example 2
A seller of enterprise software creates an ad listing for its
products, and links this listing to the topic "computer software."
The seller additionally links the listing to the user class "users
with email addresses corresponding to major software companies,"
which includes all users having email addresses ending in
microsoft.com, oracle.com, etc. The listing is thereafter displayed
primarily or exclusively on software-related web pages when such
pages are accessed by users who likely work at major software
companies.
II. KEYWORD-BASED EMBODIMENTS
The above-described architecture and process may also be used in a
system in which the listing providers bid on keywords with which to
associate their listings. As with topics, the keywords serve
generally as content descriptors that describe the type of web page
content to which the listings are to be matched. In one such
keyword-based embodiment, the listing provider may initially select
a keyword or keyword phrase with which to associate a particular
listing. Upon selecting a keyword or keyword phrase, the listing
creation application 34 looks up from the table 62, and presents to
the listing provider, a set of user classes associated with this
keyword or keyword phrase. The listing provider may then have the
option to select from this list to specify one or more target user
classes. For example, if the keyword phrase "DVD player" is
selected, the listing provider may be given the option to select
from the following list of user classes: "has purchased DVD
player," "has not purchased DVD player," "advanced electronics
user," "all users."
The matching process may be similar to that depicted in FIG. 3. For
example, when a search request is received, the matching engine 38
may initially identify a set of listings that are associated with a
keyword or keyword phrase included in the user's search query. The
matching engine 38 may then look up the user classes associated
with this keyword or keyword phrase, and if possible, select the
class (or set of classes) that best characterizes the user. If such
a class is identified, the set of listings may be further refined
by selecting only those listings that are linked to this user class
(or set of user classes). In the case of a non-search page request,
the process may be the same except that the content of the
requested page may initially be analyzed to identify a keyword or
keyword phrase that characterizes the page.
III. CONCLUSION
Although this invention has been described in terms of certain
preferred embodiments and applications, other embodiments and
applications that are apparent to those of ordinary skill in the
art, including embodiments which do not provide all of the features
and advantages set forth herein, are also within the scope of this
invention. Accordingly, the scope of the present invention is
intended to be defined only by reference to the appended claims,
which are intended to be construed without reliance on any
definitions that may be set forth in any incorporated-by-reference
materials.
* * * * *
References